Predicting breast cancer recurrence

Irvine, Calif.-based Agendia Inc. presented data from the FOCUS trial showing the clinical utility of Mammaprint, a genomic test to predict risk breast cancer recurrence, in older breast cancer patients. The population-based cohort of over 2,000 people included all consecutive breast cancer patients over 65 years of age who were diagnosed between 1997 and 2004 in the Comprehensive Cancer Center region West, the Netherlands. The aim of this study was to assess Mammaprint’s stratification of breast cancers in women over 70 years old. Results showed the 70-gene assay accurately stratified patients according to their 10-year distant recurrence-free interval, as previously demonstrated in younger patients. Moreover, even clinically high-risk patients who were classified as Mammaprint Ultralow risk did not experience recurrent disease 10 years post-diagnosis, allowing physicians and patients to make informed decisions about future treatment. The company presented the results at the 2020 San Antonio Breast Cancer Symposium.

Comparing SARS-CoV-2 PCR tests

The COVID-19 pandemic has produced a plethora of tests to detect the SARS-CoV-2 virus, but there is little data on how the performance of various assays compare. Researchers at the University of California, Los Angeles performed a retrospective analysis of more than 10,000 test results from three widely used real-time polymerase chain reaction (RT-PCR) tests for COVID-19 – CDC, Simplexa Direct (Diasorin Molecular LLC) and Taqpath (Applied Biosystems Inc.) – to assess their performance characteristics. They also retested what was left of weakly positive samples to assess sensitivity. The results showed a strong linear correlation and little bias among cycle threshold values for PCR targets, In patients with first-test negative results, 98% remained negative in subsequent testing. In retesting of weakly positive samples, Taqpath sensitivity was 97.8%, CDC was 91% and Simplexa Direct was 75.3%. “Our analysis showed no performance difference among PCR targets within the same assay, suggesting a single target is sufficient for SARS-CoV-2 detection,” the researchers wrote. “Lower respiratory tract (LRT) specimens had higher negative predictive value (NVP) (100%) than upper respiratory tract specimens (98%), highlighting the utility of testing LRT specimens when clinically indicated. NPV did not increase upon further repeat testing, providing strong evidence for discouraging unnecessary repeat testing for SARS-CoV-2.” Their analysis was published Dec. 4, 2020, in The Journal of Molecular Diagnostics.

Paying attention to the little guy in proteomics

Researchers at Stanford University have developed an AI-based method to find small open reading frames (ORFs), regions of the genome that code for proteins, through genomic analysis. Proteins that are less than 50 amino acids long, sometimes called microproteins, are important for cell-to-cell communication, and identifying such proteins and their roles especially in microbiome bacteria could give important new insights into microbiome function. However, their size makes microproteins difficult to detect via classical bioinformatics methods, and ribosome profiling requires culturing their producers, which is not always possible. The Stanford researchers used “deep learning models to improve the detection of small proteins commonly found in the human microbiome. This annotation tool is freely available along with a re-analysis of thousands of publicly available genomes,” the authors wrote. They published their paper in the Dec. 7, 2020, online issue of Cell Host & Microbe.

Increasing biomarker reproducibility

Biomarkers have shown great potential in research on neurodegenerative diseases, but many biomarker findings have been difficult to consistently reproduce. Here, a team of researchers in Sweden identifies possible causes of low reproducibility on studies on fluid biomarkers for neurodegenerative diseases, with a focus on Alzheimer’s disease, and suggests guidelines that may improve reproducibility of findings. The team looks at several manageable sources of poor reproducibility, including cohort design, preanalytical and analytical factors and statistical procedures. Assays may also react to interferents such as lipids and certain antibodies that could affect the estimation of measured concentrations of the analyte. “Assay-related factors are not only important to control when developing and launching a new assay, but also over time when using an assay in research or clinical practice,” the authors wrote. To guide the use of biomarkers, the value of novel biomarker should be compared with established state-of-the-art diagnostic techniques and shown to be reproducible in both tertiary care and primary care settings. “Despite the many obstacles and challenges in validating biomarkers in primary care, we greatly encourage this type of validation since a biomarker that passes this replication test most likely has proved to have a high level of robustness in terms of the influence from pre-analytical factors., co-morbidities and a variety of demographic factors. Such a successful validation also potentially makes the biomarker accessible for a much larger group of people.” Their work appeared in the Dec. 7, 2020, online issue of Nature Communications.